Burnout Map Reveals Which Symptoms Drive Public Health Workers Toward the Exit

Public health workers form the quiet backbone of every health system, and when they burn out, the consequences ripple far beyond the individual. A new network analysis of nearly 1,900 public health workers in China, published in BMC Public Health, has mapped precisely how the individual symptoms of burnout interlock with the desire to quit, and the results point to a small set of symptoms that may act as the engine of workforce attrition. Rather than treating burnout as a single undifferentiated feeling, the study treats it as a system of interacting components, some of which matter far more than others when it comes to whether a worker starts scanning the job market.

The research team, led by Guimei Chen, Songjie Wu and Xiaoxia Zhang, with Jie Ke and Ke Liang as corresponding authors, conducted a nationwide cross-sectional online survey between October and November 2024. A total of 1,887 public health workers from across China completed the survey. The sample was predominantly male, with 59.1 percent of respondents men, and 59.4 percent were older than 30 years. Just over half, 51.8 percent, were public health physicians, the professional group that carries much of the day-to-day burden of disease surveillance, prevention and health education in China’s public health infrastructure.

To measure burnout, the researchers used the Maslach Burnout Inventory, or MBI, the most widely used instrument in occupational health psychology. The MBI decomposes burnout into three dimensions: emotional exhaustion, the feeling of being drained by work; depersonalization, a cynical detachment from the people the work is meant to serve; and reduced personal accomplishment, a diminished sense of competence and meaning. Intent to leave was assessed with O’Reilly’s four-item scale, a compact measure of how seriously a worker is contemplating departure from their current position.

The methodological innovation of the study lies in its analytical approach. Instead of running a traditional regression that estimates the average effect of a burnout score on quitting intentions, the team estimated a regularized partial correlation network. In this framework, each burnout symptom and each intent-to-leave item is a node, and the statistical associations between them, after controlling for all other variables in the network, are the edges. The result is a map, not unlike a subway diagram, in which the busiest stations and the most heavily trafficked lines can be identified. Centrality indices quantify which nodes are most strongly connected to the rest of the network, while bridge centrality indices identify which symptoms connect otherwise separate communities of nodes, in this case the burnout cluster and the intent-to-leave cluster.

The network revealed dense interconnections between burnout symptoms and intent to leave, confirming that the two constructs are not merely correlated in the aggregate but are woven together at the level of individual symptoms. Three symptoms emerged as the most central in the entire network. The first was severe exhaustion, captured by item MBI5 of the Maslach Burnout Inventory. The second was declining work interest, item MBI6. The third was doubts about the value of one’s work, item MBI8. In network terms, these three nodes sat at the heart of the web, meaning they maintained the strongest direct associations with the surrounding symptoms even after the regularized estimation pruned weaker connections.

Perhaps the most striking finding concerned the bridge symptoms. Two items, MBI6 and MBI8, both of which belong to the depersonalization dimension of burnout, showed the strongest bridge centrality linking depersonalization to intent to leave. In plain terms, when a public health worker begins to lose interest in the work and to question whether the work has value at all, those two experiences appear to be the primary conduits through which burnout translates into a concrete intention to quit. Emotional exhaustion, by contrast, while highly central within the burnout community, did not serve as the strongest bridge to leaving, suggesting that feeling drained and wanting to leave are connected but not identical phenomena.

This distinction has real practical significance. Traditional burnout interventions often target exhaustion, on the assumption that reducing fatigue will reduce turnover. The network evidence suggests a more nuanced picture: exhaustion is a core symptom of the burnout syndrome, but the pathway from burnout to intent to leave appears to run primarily through the depersonalization symptoms of declining interest and doubts about work value. If the network structure reflects causal dynamics, as network theorists propose, then interventions that rekindle a sense of meaning and engagement in the work itself might do more to retain workers than interventions focused solely on workload reduction, although the authors are careful to frame these as candidate targets rather than proven causal levers.

The researchers took seriously the question of whether their network map was statistically trustworthy. Using bootstrap methods, they evaluated the stability and accuracy of the network across all indices, and the network demonstrated good stability and accuracy. This matters because network analyses can be sensitive to sample fluctuations; a network whose edge weights and centrality rankings shift dramatically under resampling would offer little guidance for intervention design. The reported stability indicates that the identification of severe exhaustion, declining work interest and doubts about work value as central nodes, and of MBI6 and MBI8 as bridge symptoms, is unlikely to be a statistical artifact of this particular sample.

The context of the study is important. The COVID-19 pandemic exacerbated burnout and intent to leave among public health workers worldwide, and China’s public health workforce was no exception, having borne an extraordinary burden during pandemic control efforts. Understanding which psychological states most strongly propel workers toward the exit is therefore not an academic exercise but a workforce security question. Attrition among public health physicians and allied professionals erodes institutional memory, weakens surveillance capacity and increases the load on those who remain, potentially feeding a self-reinforcing cycle of burnout and departure.

The authors conclude that the three central symptoms, severe exhaustion, declining work interest and doubts about work value, together with the bridging role of the two depersonalization symptoms, highlight specific symptom targets for intervention strategies aimed at reducing burnout-related intent to leave among public health workers. The study received no external funding, was approved by the Medical Ethics Committee of the Second Hospital of Changsha, and obtained electronic informed consent from all participants. As a cross-sectional survey, it captures associations at a single point in time and cannot by itself establish that treating a given symptom will prevent resignation; longitudinal and intervention studies will be needed to test that proposition. But by replacing the blunt question of whether burnout predicts quitting with the sharper question of which symptoms carry the strongest connections, the study offers workforce planners a more precise map of where to intervene, and it suggests that restoring meaning and interest in public health work may be as important as reducing the hours that drain it.

Subject of Research: Network analysis of burnout symptoms and intent to leave among Chinese public health workers

Article Title: Mapping the association between burnout and intent to leave: a network analysis of public health workers in China

Article References: Chen, G., Wu, S., Zhang, X., Xie, H., Zhu, S., Wan, L., Zou, S., Liu, J., Ke, J., & Liang, K. (2026). Mapping the association between burnout and intent to leave: a network analysis of public health workers in China. BMC Public Health. https://doi.org/10.1186/s12889-026-29427-1

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29427-1

Keywords: burnout, intent to leave, public health workers, network analysis, Maslach Burnout Inventory, depersonalization, emotional exhaustion, occupational health psychology, China, workforce attrition, BMC Public Health, cross-sectional survey